I am interested in the neural basis and the computational principles underlying human decision making. My research focuses on value-based decisions and aims at understanding how the values of available choices are computed from perceptual evidence, internal preferences and the subjective model of the choices context. My current work involves simultaneous measurements of electrical and hemodynamic activity (EEG/fMRI) in the human brain to identify the spatio-temporal characteristics of the evidence accumulation process in value-based decisions.
I am not quite sure when I happened to start thinking about what’s happening when thinking but I soon realised that whatever it was, it was happening, physically, behind our eyes. This is why I chose to study Physics in La Sapienza University in Rome where, in 2009, I graduated with a thesis on a neural network model of sensorimotor control.
At that point it was clear to me how fundamental it was approaching the study of the brain through a principled modelling of the computations it performs. This is why on September 2009 I came to London to study Computational Statistics and Machine Learning at UCL. One year later I graduated with a project on a statistical model of attention based on approximate inference, which Microsoft judged to be the best of my course, that year.
In November 2014 I obtained a PhD in Neuroscience at the UCL Institute of Ophthalmology, in the laboratory of Prof.Carandini. Under his supervision, I used Wide-field Intrinsic and Fluorescence Imaging to study the relationship between hemodynamic and neural activity in the mouse visual cortex. As a part of my PhD I took the Theoretical Neuroscience course at the Gatsby Computational Neuroscience Unit, which was also, hopefully, the last exam of my life, and I was quite lucky to publish twice as first-author in leading international peer-reviewed journals of my research field.
The first paper, entitled “Fast hemodynamic responses in the visual cortex of the awake mouse” (Journal of Neuroscience, 2013) studied the effect of brain state on neurovascular coupling, the mechanism at the basis of the most commonly used neuroimaging techniques (i.e. fMRI).
The second paper, entitled “Local and global contributions to hemodynamic activity in mouse cortex“ (Journal of Neurophysiology, 2016) expanded the scope of my first publication by finding more profound effect of brain state on vascular activity in the mouse cortex suggesting a novel method to analyse neuroimaging data to reveal the differential contributions of external stimuli and internal dynamics on cortical hemodynamic activity. The image on the right, showing a strong correlation between pupil dilations and cortical hemodynamic activity, gained the paper a robust coverage on social media, ending up in the top 5% of all research output scored by Altmetric. If you you want to hear me chatting about this result with my two scientific heroes, you can do that here.